Content
What can you learn from YouTube without the API?
Every YouTube tool shows search volume and a competition score. Where those come from, what the comments are genuinely worth, and which numbers are synthesised.
How MarketWar handles it
Without a connected data source, every volume, competition, opportunity and score figure here is synthesised from a hash of the input and labelled an ESTIMATE ONLY, with an explicit disclaimer in every payload: it never claims real YouTube view counts, subscriber figures or ranking data. Sentiment and pain-point clusters are lexical heuristics over the text you supply, not audience data — so what you get back about your comments is a reading of YOUR comments and nothing more.
YouTube is the second-largest search engine and the one small businesses most often ignore, usually because the tooling around it is the least honest in marketing.
The numbers nobody has
YouTube does not publish per-keyword search volume. So every tool showing you "9,400 searches a month" for a phrase has modelled it, and the models disagree with each other — exactly as they do for web search.
With no data source connected, the figures here are synthesised from a hash of the input and labelled estimate only, with an explicit disclaimer travelling in every payload. The engine does not claim real view counts, subscriber figures or ranking data, because it does not have them.
That is a less impressive screen than the competition offers. It is the difference between a scaffold and a lie.
What is genuinely valuable, and it is free
The comments. Not yours necessarily — the comments on the best videos in your subject, including competitors'.
People use comments to ask the questions the video did not answer. Those questions are:
- a list of what to make next, written by the audience,
- phrased the way real people phrase it, which is what you want in a title,
- evidence of what the existing videos are failing to cover.
For a trade, a clinic, a consultancy — anybody whose customers research before buying — twenty minutes reading comments on the top five videos in your subject is worth more than any keyword tool, and it costs nothing.
The engine clusters them into pain points, and the rule is the same as everywhere else here: lexical heuristics over the text you supply, labelled as estimates, not audience data. It reads your comments and reports on your comments.
Titles and thumbnails: patterns, not predictions
What the successful videos in a subject have in common — structure rather than a score. A pattern you can see is worth more than a number you cannot interrogate, because you can look at it and decide whether it applies to you.
The honest short-form workflow
Record one proper answer to one real customer question. Do not script it tightly; say it the way you would on the phone.
Then pick which thirty seconds is worth posting — and note that the clip that sells and the clip that gets watched are usually different moments, which is why those are scored separately.
Caption everything. Most viewing is silent, and an uncaptioned clip whose selling point is spoken sells nothing.
What this cannot tell you
What will get views. Nothing can, including the tools that publish a confident number for it. Views are decided by a recommendation system optimising for its own retention metrics against everything else published that day.
What you can control is whether the thing you made answers a question somebody actually asked. That is why the comments matter more than the scores.
The question to ask any YouTube tool
Which of these numbers came from YouTube?
If the honest answer is none of them, the tool can still be useful — for structure, for patterns, for reading your own comments at scale. It just cannot tell you what to expect, and a tool that implies otherwise is selling the one thing it does not have.
What this does not do
It cannot tell you what will get views. Nothing can, including the tools that publish a number for it, and the part of this that is genuinely useful — mining your own comments — needs you to paste the comments in.
Common questions
Where does YouTube keyword volume come from?
Not from YouTube, which does not publish it. Every tool models it, and with no source connected ours says plainly that the figure is synthesised rather than presenting it as data.
Then what is actually worth doing?
Mining the comments on videos in your subject — yours and other people's. The questions people repeat are a free, specific list of what to make next, and they are real text rather than a modelled score.
Is the sentiment analysis reliable?
It is a lexical heuristic over the text you supply, labelled as an estimate. Useful for sorting a large pile; not a measurement of how an audience feels.
What about thumbnails and titles?
Pattern analysis of what the successful videos in a subject have in common — structure rather than a prediction. Worth more than a score, because you can see the pattern and judge it yourself.
Related
Which thirty seconds of your video is worth posting?
Clip tools find the interesting moment, which is not the same as the selling one. Eight separate scores instead of one, and the dimension that returns nothing.
Where do keyword volume numbers actually come from?
Every SEO tool shows a monthly search volume and a difficulty score. What is behind those, why they disagree with each other, and what ours says about itself.
What are your customers actually telling you?
Reviews, surveys, support messages and sales calls say the same three things, and nobody reads them. How clustering differs from inventing.